Hi,
A couple of months ago we had the discussion why the agent is currently
doing a "df" on the mountpoint of a NFS pool [0]
Edison told me this was due to a bug in libvirt under RHEL6.
Today I tested with RHEL6 and CentOS6 with a couple of NFS mounts and it
works fine.
After issuing a pool-refresh libvirt will reply with the correct
allocation and capacity of a pool.
I'm thinking about removing this code from the agent and fully relying
on libvirt for this, any objections or comments?
I've tested with libvirt 0.9.4, which is default under RHEL6.
Wido
[0]:
https://git-wip-us.apache.org/repos/asf?p=incubator-cloudstack.git;a=blob;f=agent/src/com/cloud/agent/storage/LibvirtStorageAdaptor.java;h=6ee69847d204bd40499ed0adad65f29d87a93164;hb=master#l391
